Seth Goldstein

Automatic Motion and the Gestures of Social Media

The Attention Economy is based on human behavior, behavior which includes the full range of physical gestures that we use to express our inner intentions.  These gestures could be saying something,  stretching our arms, walking past a group of people, or reading a book.

Many gestures end in themselves, as discrete actions.  Other gestures trigger subsequent gestures: for the person who performs the gesture (such as throwing a ball into the air and then catching it on the way down) or for the one that receives the gesture (such as calling somebody on the phone who then picks up their phone).  Since its introduction as mass consumer platform over ten years ago, the Internet has attracted an increasing amount of the latter: electronic gestures that use keyboards and mice (and increasing voice and video) to trigger subsequent gestures from others across the Internet.  It is not simply that more human behavior is being expressed through the Internet; entirely new kinds of human behavior are being created by virtue of the medium.  Many of these gestures are able to authenticate intentions with great subtlety so as to increase the ratio of signal to noise.

These are the gestures of social media, where influence is measured by the amount of Attention one gets relative to the amount of information one gives  The most influential online individuals combine a high pagerank that lands them above the Google fold (ie their unique algorithm) and a rich personal stream of syndicatable "what makes me me"-ness:  ie their natural API. 

The Turing test measures whether a computational response is mistakable for an actual human being.  George Dyson’s recent essay Turing’s Cathedral imagined this test to be central to Brin and Page’s agenda for Google.  They are trying to build an artificial intelligence whose automatic responses to queries seem natural and humanistic.  In order to experience the rich social media of a post-Web 2.0 landscape, however, we need to venture beyond the shadow of the Google page-rank Golem.  How can we trust an electronic reality based solely on the single perspective of pagerank (ie the link model), when our actual behavior seems more like, in the words of William James, a “blooming, buzzing confusion” of human electronic gestures.

"Egosystems, not Ecosystems"

Stallman, Torvalds, Doc Searls, R0ml and many others have established the underlying value of an Open Source ecosystem.  At the first public meeting of AttentionTrust.org in October 2005, Tim O’Reilly suggested that the next generation of the Web (3.0?) was going to be about "Data Inside".  And so if you apply the discipline of a truly Open system to all of the behavioral data that is expressed by these electronic gestures, you end up with a new organizing framework- a framework of egosystems which, for the first time, bind human behavior to the transport policies of Open Source ecosystems.  My wife Tina Sharkey came up with this brilliant expression, "egosystems, not ecosystems." in which the notion of an egosystem can be used to express the rich variety of our interactions across the Internet:  email, search, browsing, buying, tagging, chatting, and other electronic activities.  One egosystem may interact with other egosystems in infinitely different ways.  And so even at maximum resolution, this social media fabric appears richly woven and densely textured:  tiny threads of gestribution bundled transparently around eachother.

Enter the worms (which happen to be the subject of the last section of the last chapter on Arbitrage).  These little creatures that both consume and produce in equal measures provide the best model for understanding the physics of the Attention Economy.

Hamlet explains that:

We fat all creatures else to fat us, 
and we fat ourselves for maggots. Your fat king and your lean beggar is
but variable service- two dishes, but to one table.

So Hamlet 2.0 might say:

We encourage others to participate so that we may consume them
and we make ourselves interesting for the blogosphere.  Your Internet CEO and your Joe Blogger are just different algorithms- two APIs, but to one network.

What distinguishes social media from other forms of media is this worm-like behavior, where production and consumption occur simultaneously.  You can look at it as a form of therapy, where individual data is constantly being reorganized (quantitatively) to generate richer (qualititative) meaning.  This is the  Gillmor moment, when the naval-gazing gesture of an individual’s Attention feeds back, anonymously, into crystal-clear affinity pools of metadata.  Inference and influence reverberate off of eachother, turning a mess of discrete gestures into a continuous Gesture stream.

Attention Science

Attention has forever been described as an absence in terms of distractions, deficits and obligations.  Media Futures is a model for talking about Attention as something more; as an active substance that we create, express, share and remember.  This Attention substance- let’s call them attentrons- operates at a size and frequency that makes it invisible to our naked eye.  Fortunately, none of us need travel very far to reach the Attention super collider, where we can test for traces of these attentrons.  The mashups that emerge on Digg are powerful reactions of data and people.  From the new dance clubs in Second Life ripple new patterns of social networks.  How can we analyze these seemingly spontaneous phenomena so as to make visible the elements and interactions that drive their behavior?

This is the project of Media Futures: to come up with a conceptual
schema that synthesizes the  constructivist aspects of social media
and the analytical rigor of the Attention Economy.  It is a
language game:  as complex as Chess, and as fleeting as Chutes and
Ladders.  Computer scientists in the 1950’s gave us the building blocks of Input, Store, View and Output;  MS-DOS gave us operations like copy, file, print and run; Unix gave us root access.  In the Attention OS, it’s the 5 A’s: Automata, Algorithm, API, Alchemy and Arbitrage.

Maggie Dillon has done a great job aggregating brief histories of each topic, both to establish its historical credibility and to expose the subtleties of its use over time.  This syntax of Media Futures is valuable, however, only in so far as it is applied to real social media use cases.  These new services require active minds, creative imaginations and lots of quality code.  What’s missing is the sense of continuity between early computing history and the widget design for next week’s release.  And that is what we are working on here.  Enjoy.

Next: History of Automata

Foreward

Since writing a series of essays on Media Futures in the Spring of 2005, I have spent the last year or so investing in and building out various data services.  These include: a lead generation marketplace at rootexchange.com, whose first vertical is mortgage;  a clickstream media platform at root.net, the command line for a new Attention-based OS; AttentionTrust and the promotion of its four principles of property, mobility, economy, and transparency (AttentionTrust.org is now the #2 organic search result for Attention on Google); the “crystallized attention” (tag) company del.icio.us, which was acquired by Yahoo!; and finally, Majestic Research, the investment firm that uses online consumer behavior for its equity models and which I co-founded in 2002.  Majestic is the name I used for this blog on typepad, and its subtitle transparent bundles was an attempt to describe how investment research and trading should operate. 

Sometimes it feels like I am working on a number of disconnected activities.  But enough of the time it feels like they are all connected in a deeper kind of way.  They all deal with consumer Internet usage; and more specifically, they share the common goal of maximizing the signal-to-noise ratio of online data in order to elicit the highest fidelity copy of an individual’s Attention.  This is not an easy problem to solve, as the interface between one’s mental focus and the TCP/IP protocol is indirect at best.  What we have as proxies are clicks, searches, tags, forms and other types of user generated media.  From the interplay of these artifacts we– as in the royal Web 2.0 we– are busy coding a social media fabric, the center of which always seems but one release away.

Unlike most media properties, Attention is inherently unstable and indeterminate.  Describing Attention is like making a movie inside of a house of mirrors, where it is impossible to keep the camera itself out of the picture.  It is because of this Heisenberg-like uncertainty principal that passive behavioral data provides the better indicator of pure Attention than explicit user generated content such as ratings, reviews and tags (which change the substance of Attention as they reflect it).  As we review the history of Attention, it seems always caught in its own shadow; artists and actors want Attention and create works and performances to “attract” and “capture” it.  Only recently have certain of us (guided by Goldhaber’s theories on the matter) come to see Attention in its own light: as a material substance that moves from one human being to another like a language or a liquid.  Our cognitive framework for Attention needs to shift from metaphors of coercion to metaphors of creation.

The distributor of Attention may indeed be influenced by that receiver who provides the most interesting information, but still the former maintains control over who gets his Attention.  It is this choice the individual has over where he spends his Attention that underlies the theory of Media Futures.  This new organon assumes that the user is in control of the media that he and his network of social and commercial relationships create.   With the traditional consumer now in control over the means of social media production, the traditional media company now needs a new value-creation model– one based on consuming the most relevant electronic gestures of its audience, rather than one based on producing the most engaging content.

For a broader dialectical context, I would encourage you to  tune into the following writers

• Steve Gillmor
• Michael Goldhaber
• John Hagel
• Umair Haque
• Doc Searls
• AttentionTrust.org

Attention:  The underlying instrument of Media Futures

“Attention is scarce,” Michael Goldhaber writes, “because each of us only has so much of it to give, and it can only come from us – not machines, computers or anywhere else.”  It is in cyberspace, he argues, that a new type of economy comes into its own: this is the attention economy, an economy based on what is both “most desirable and ultimately most scarce.” 

Goldhaber’s principles of the attention economy enter into a long-standing dialogue among art historians and cultural theorists about the techniques and implications of attention in the production and reception of media.  As art historian Michael Fried argues in Absorption and Theatricality, it was first in the writings of Diderot that the terms of attention assumed critical in addition to rhetorical significance.  A painting, Fried writes, “had first to attract (attirer, appeller) and then to arrest (arrêter) and finally to enthrall (attacher) the beholder, that is, a painting had to call someone, bring him to a halt in front of itself, and hold him there as if spellbound and unable to move.”  Then, it was the media itself being consumed that did the work advertising does today: it was up to the media itself to call out to consumers for their attention.

The Beauty Salon

Of course, in today’s salons, we are more likely to consume the news of celebrity hook-ups than the spectacle of high art: that the salon is still a place for to see and be seen is telling.  In the eighteenth century, the salon was a privileged site for the bourgeoisie to consume, contemplate and discuss art and literature – truly a place for seeing and being seen.  We pay visits to an entirely different type of salon today: we go in preparation for – or to increase our chances of – the condition of being seen.  By doing work on our bodies – by taking clippers to our dead cells, by taking tweezers to our brows, we might too do our own advertising: we might attract, arrest and enthrall the passers-by.  We pay to increase our chances of being beheld – consumed, contemplated, discussed; we pay so that others might pay attention to us.

This is the to be seen half – but that which we see in salons, besides other guests questing to improve their own appearances, is the set of people important enough to be seen by the masses: celebrity.  Magazines like People and Us Weekly, which adorn the waiting areas, promise a behind-the-scenes look at the lives of people who have entertained us on stage or on the big screen, or perhaps even written books for our edification or delight.  These celebrities have captured the public’s attention with their work, and they certainly capture the public’s attention with their play.  And the placement of celebrity magazines in salons suggests the possibility that by altering our appearance, perhaps in the fashion of the star du jour, we might capture more attention.  At the end of this line of fantasy is the possibility of our own presence in such a magazine, the possibility that the banalities of our own lives will be represented in the world of others and put out for consumption by third, fourth, millionth parties.  We will be worthy of attention.

The Internet Salon

The truth is that our own private gestures are constantly being recognized, represented and put out for public consumption – and in real time.  Moreover, these gestures are at the same time being plugged into calculations to predict our future behavior, calculations which promise a personalized experience to those who click (and profit to those who calculate).  The playing-out of these phenomena takes place, of course, on the Internet.  This is an economy of attention – one, as Goldhaber argues, that is different from any economy seen before: “In its pure form, it doesn’t involve any sort of money, nor a market or anything closely resembling one.  It involves a quite different pattern of life than the routine-based, industrial one…What matters is seeking, obtaining and paying attention.”  The economy’s “characteristic form of property” is “the attention that is readily available to its ‘owner’ from other people, which depends on what attention this owner has gotten in the past”; it is a property “located, quite literally, in ‘the minds of the beholders.”

In his work on the attention economy, Goldhaber views the movement toward cyberspace as analogous to the move of western European civilization to the New World of the Americas around the time of the birth of the market economy.  “Unimpeded by the remains of feudalism,” he writes, “the market-industrial system in fact took most complete hold here in North America first.  From here, much later, it swept back to complete its conquest of the western European motherland, along with the rest of the globe.”  Similarly, “Cyberspace will be the ‘place’ where the new economy moves ahead most dynamically, but the strength gained in the process will eventually sweep back to dominate the rest of life.”

If Attention is indeed the substance of focus (that which registers our interests by indicating our choice for certain things and choice against other things), then Internet is the most fertile ground for the development of the Attention Economy;  for the Internet (and particularly web services) allows the recording and sharing of our choices, of our Attention, in real-time.  These choices of ours are manifested by the binary gestures of the keyboard and mouse.  With each click, our own narratives expand.  With each move to create a tag or a link, our narratives expand.  With each search, with each subscription, our narratives expand to tell the story of which team we follow, where we will be taking our next vacation, which conference we are planning to attend.  The gestures of our lives are recorded, and we become represented – on “Top 100” lists, blogrolls and Flickr badges of different  sizes.  And the narratives of our electronic Attention gestures have even crossed back into offline mass media: on CNN’s headline news or American Idol’s SMS voting.  We may not be followed by paparazzi, but airtime on national television is a start.   

The sociological, psychological and economic forces at play in this discussion warrant extended research.  As such, it is a daunting task to wrestle with the history of social media and probe into its future development.  The Internet is a dynamic site of all sorts of production and consumption, a place where familiar models are broken and reinvented, a place where the material being consumed is dynamic, produced on the fly.  We have tags, wikis, social networks and other forms of social media – we have new forms of media being created by everyman for everyman, and at any time, in any place.  And works in these media are being created at a far higher rate than they are being consumed.  Power and value shift, become redefined; the very possibilities of our personhood shift, become redefined.
The more we express ourselves electronically, the more residue we leave behind in this ever-growing, ever-changing landscape – shadows of our digital actions scattered about held together not by gravity, biology, optics but by algorithms and APIs.  The economics of behavioral data, and the electronic media gestures that constitute this data, reveal themselves in an analysis of Attention.   This is the goal of updating Media Futures one year later:   Over the coming weeks, I will write the five-boned skeleton of A’s into the skin of Attention: 

It is a body of work that seeks to better understand our gestures in social media, the very articulations of our attention and intentions – a pyramid topped by Attention and flanked by:

• Automata- Human inspiration
• Algorithm- Patterns of behavior
• API- Natural expression
• Alchemy- Value creation
• Arbitrage- Economic discovery

This is a model I see as most compelling in examining the delta of change, the fertile crescent lying between Wall Street and Madison Avenue.

Note:  I am fortunate to be working with an extremely thoughtful and lyrical research assistant in Maggie Dillon, who recently graduated Princeton and who will be studying art history and media theory next year at the University of Cologne.  She likes to refer to herself as a "an aspiring poet and brewer from the country’s heartland," which is clearly the kind of midwestern pragmatic spirit that this blog needs more of!

QUESTIONS AND ANSWERS

Q:  It’s not easy to become an insurance company but it’s easy to become a broker.  So the question is, have you guys attacked any of that piece of it, not just the originations, but also the insurances that accompany them?

Ranieri:  There is nothing that says in the same way you get his lead from online you cannot deliver all this other kinds of information online, which is normally difficult in a traditional process.  Unless the consumer proactively tries to break out of the corral, it doesn’t happen.  Here he’s already out of the corral; he’s online.  You follow?  And I think what you’re going to see is a lead may become simply a sequence of leads.  The lead won’t simply be a lead for a mortgage.  You might sell the lead four times to four mortgage companies; but you may sell that same lead beyond that, because as sure as God made little green apples, there will be a sequence of insurance companies or title companies who want that same information.

Q: This is a question about the Exchange that you have built.  As you and I both know going way back in the lead generation business, it’s a supply driven business so if you can’t get the supply then there’s nothing to exchange.  So even though I agree with Lew that this could potentially be a big win for mortgage brokers because they’re getting more leads, better leads, a price based on the real value … even though I believe it could be a win for the consumer because probably they’re getting well-priced goods and services, I don’t see any reason in the world why any of the companies out there that are generating real estate or mortgage leads would want to cooperate.

Goldstein:  We call them aggregators, and they are generating supply across the industry:  LowerMyBills, Nextag, LendingTree, LoanWeb, Adteractive, there’s a number of them.  Many are quite good and are generating high quality mortgage leads that lenders are purchasing.  The Exchange is open and available to them at minimal cost as an opportunity to discover price and to clear inventory that only has 24 hours before it goes stale.  Some of them may have their own direct relationships where they’ll sell the lead once or twice, but for the third or fourth time they might use the Exchange.  Most of these aggregators purchase their media inventory from publishers.  A consumer of one of these aggregators is typically coming in through a search engine or a banner.  The natural suppliers of this inventory are the newspapers, the search engines, the media companies, all of whom up until now haven’t really had a channel to sell leads directly through.  They have had to sell impressions and trust that they are getting a fair price for their impressions relative to the value of the underlying leads they could be generating.  Many publishers are indeed getting fair prices, but there are some publishers who have relevant search terms or relevant content areas that might monetize better it they were able to go directly to our Exchange.  And so we are committed to enabling both traditional publishers who are capturing the direct attention of the consumer and aggregators who, in most cases, are purchasing media to generate leads. 

Ranieri: In the regular, non-refi mortgage business, that volume is not controlled by an aggregator; as an example, it’s controlled by realtors.   Realtors turn those leads into money simply by, at this point, becoming a mortgage company to convert the lead into their own servicing.  I think I could argue if this process is as efficient and powerful as I believe it will be, they will be a lot better off just accessing our Exchange.  It will be much more cost effective.  Every homebuilder in the country decided they had to become a mortgage company to capture that portion of the revenue.  Our Exchange will give them the ability to exit a business they really never wanted to be in.  They just want to keep that portion of the income that came from the result of building the house.  This gives them a much more cost efficient way of doing that.

Q:  That’s why I’m saying it’s great for the demand side but maybe it’s not so great for the supply side.  If I’m a publisher and I’m making $10,000 a month from an aggregator, why should I drop them and sell leads directly to your Exchange?  I think you’re going to get some resistance.

Ranieri: One of the largest mortgage companies in America at the beginning of all this was Loomis and Nettleton, and they were a monopoly.  They viewed the advent of the mortgages security market as a risk, as something that was competition because in those days mortgage companies were brokers.  They weren’t what mortgage companies are today.  They actually simply brokered a loan between thrifts or between thrifts and pension funds.  And here I was suggesting that mortgage companies become originators:  get into the mortgage business themselves, take the mortgages they create and turn it into a security and sell it.   So that their whole business wouldn’t be going to the Arizona Banking Conference and lining up four new thrifts to sell them into.  But I was preaching heresy.  And this great, giant company, Loomis and Nettleton decided “a pox on my house” and that they weren’t going to change.  And unfortunately it turned out for them that I wasn’t all wrong.  And there’s no longer a Loomis and Nettleton. 

Goldstein: But there were plenty who adapted to the evolution.

Ranieri: It was actually good for mortgage companies.  In fact, if it was bad for anybody it was bad for the thrift who wanted to show up at 10 and go golfing by 3.  The mortgage companies inherited the earth. 

Q: Your comments about refinance product were focused largely on refinancing to a fixed rate product, any thoughts on the option ARM product?

Ranieri:  We’re talking about a very sensitive topic so I’ll just take it on head-on.  Option ARMs in their original design were created for middle income, upper-middle income people who understood the nature of the risk in the option arm and for whom it worked very well because they could manage the risk.  They could de-lever if they so chose.  That was the issue; you could de-lever.  When the option ARM started to be used as a vehicle to qualify the unqualified, it became an abuse, in my opinion.  Actually, the first arm was called a floater.  The first ARM structure wasn’t called an ARM, it was called a floater and the rest of us called them sinkers because in the next cycle they never ever hit par again.  So an option arm in its most aggressive form where you have a 1%, then it rolls to a 1% initial for 90 days, 4% and then it rolls to fully indexed after the year.  And you can choose to stay at 1 percent.  Think about it.  That’s an 85% loan to value and you choose to stay at 1%. In less than two years you’ll be at 115% loan to value and you can’t go anywhere.  I’ll write in blood for any of you who want to have this bet with me.  I will tell you there’s a 100% correlation between equity and foreclosure.  Nobody lets you take a house that he has net equity in and nobody tries to stop you when he has negative net equity.  So you drive a loan up to 115% in two years, unless housing’s appreciating to offset that, you are going to have an unhappy amount of delinquency and loss on foreclosure.  So I think the option arms are a very useful product as long as you don’t use them for what they were never designed for.

Q: Servicing value?

Ranieri:  Servicing value will follow, depending on who’s buying the servicing.  Thrifts were for a very long time very naïve about the risks to certain kinds of consumer and option ARMs and so the option ARMs were 5/8’s of a point regardless of whether they were upper-middle class or being used as a way to qualify the unqualified.  Then the FDIC, the regulators, the federal financial oversight group came out with guidelines and said to financial institutions, banks, credit unions and council and state insurance companies: “If you do this, I have news for you.  You can do it.  We won’t stop you.  It’s not technically illegal, but when we come in, we’re going to ask you to do a series of algorithms and if you cap out in two years we’re going to make you write these things down; we’re going to make you take reserves against it.” And so the aggressive AM arm loans now are trading at pretty substantial discounts to the 5/8s and 3/4s of a point based on those regulatory algorithms.

(Thread 6/6 of RootExchange/s speaker series: #2 June 13, 2006 with Seth Goldstein interviewing Lew Ranieri at NY office of Root Markets)